Hyundai, AIM Intelligence, and LG AI Research leverage NVIDIA OpenShell and Nemotron to enhance AI agent security and efficiency in South Korea.
South Korea's AI ecosystem is making strides in secure and efficient AI agent development through collaborations with NVIDIA. Key players such as Hyundai Motor Group, AIM Intelligence, and LG AI Research are leveraging NVIDIA's OpenShell runtime and Nemotron datasets to enhance AI agent capabilities and security.
Hyundai Motor Group is exploring NVIDIA NemoClaw blueprints and the OpenShell runtime to support secure AI agents in enterprise workflows. Meanwhile, AIM Intelligence, a Korean AI security company, is integrating its Starfort platform with NVIDIA OpenShell. This integration aims to combine OpenShell’s sandboxed execution and policy-based controls with Starfort’s capabilities in detecting malicious instructions and preventing data leakage. AIM Intelligence is also considering a proof of concept for the public energy sector, with potential applications across industrial and robotics environments.
LG AI Research has introduced API access for its K-EXAONE-2.5 model, which was developed using NVIDIA Nemotron datasets, including Nemotron-Personas-Korea. This dataset comprises 7 million synthetic Korean personas grounded in official Korean statistics, enabling developers to create AI models tailored to Korean contexts. NVIDIA Nemotron provides a suite of open models, datasets, and techniques designed to help developers build specialized and efficient AI agents.
These advancements align with South Korea's broader AI strategy, which includes partnerships between NVIDIA and institutions like KAIST and Seoul National University. These collaborations focus on building national AI factories and advancing agentic AI research. NVIDIA's ongoing partnerships with companies such as Naver and SK Telecom further underscore the nation's commitment to becoming a global leader in AI innovation.
Source link







